Solution Characterization, Conformation Analysis, and Dynamics of

Pauline M. N. Low, Yuk Lin Yong, Yaw Kai Yan, and T. S. Andy Hor, Sik-Lok Lam, Kam Kwong Chan, Chi Wu, and Steve C. F. Au-Yeung, Yuh-Sheng Wen and ...
0 downloads 0 Views 635KB Size
Inorg. Chem. 1994, 33, 2407-2412

2407

Solution Characterization, Conformation Analysis, and Dynamics of the Ligand l,l'-Bis(dipheny1phosphino)ferrocene (dppf) in Re~(p-OMe)2(p-dppf) (CO)6 Sik-Lok Lam, Yu-Xin Cui,t and Steve C. F. Au-Yeung' Department of Chemistry, T h e Chinese University of Hong Kong, Shatin, New Territories, Hong Kong

Yaw-Kai Yan and T. S. Andy Hor' Department of Chemistry, Faculty of Science, National University of Singapore, Kent Ridge 051 1, Singapore Received November 30, 19930

The dynamic process involving the diphenyls (4, 4') in 1,l'-bis(dipheny1phosphino)ferrocene (dppf) in [Re2(pOMe)2(p-dppf)(CO)6] at 228 K is characterized by conformational analysis. At this temperature, the fluxionality of the dppf ligand ceases whereas the dynamics of the phenyl rings about the P-Ci,(cp,cpr) (the "ipso" in the phenyl rings) was rapid on the N M R time scale down to -85 'C. The fluxional barrier (AG*cxp= 46.4 f 1.O kJ/mol) for the interconversion of the R and the S forms was determined by variable-temperature 1H N M R spectroscopic studies. The range of the Cp proton chemical shifts varies from 6(lH) = 3.00 to 5.68 ppm and is the largest reported. The [Rez(p-OMe)z] core was found to bestatic throughout the temperature-lowering process indicating that fluxionality is thus confined to the dppf fragment in the molecule. Ring current effects similar to those observed for the internal protons (Hi) in cyclophane molecules are attributed to be responsible for the anomalous proton shifts. Molecular modeling calculations indicates complete ring flipping of the diphenyls about their respective P-Ci, bonds was not possible because of the steric congestion about the ferrocene. The torsional twist of the diphenyls about their P-Ci, bonds was correlated, restricted, and in opposite directions. The ranges for the torsional angles 4; and 4t were determined respectively to be 92.6" 54; I112.4' and 147.6' I4t I177.8'.

Introduction Ferrocene-based diphosphine complexes such as 1,l'-bis(dipheny1phosphino)ferrocene (dppf) have been shown to be useful ligands because of their tunable stereogeometries, variable coordination modes,'-8 and the electrochemical9 and catalytic1S20 activities brought to their complexes. The skeletal flexibility of dppf is of particular importance as it represents the single most

* Authors to whom correspondence should be addressed.

t On leave from the Department of Chemistry, Lanzhou University, Gansu,

P. R. China. 0 Abstract published in Advance ACS Abstracts, April 15, 1994. Hor, T. S. A.; Phang, L.-T. J . Organomet. Chem. 1989,373,319-324. 1 Haggerty, B. S.; Houscroft, C. E.; Rheingold, A. L.; Shaykh, B. A. M. J . Chem. SOC., Dalton Trans. 1991, 2175-2184. ' Baker, P. K.; Fraser, S.G.; Harding, P. Inorg. Chim. Acta 1986, 116, LS-L6. Baker, P. K.; van Kampen, M.; ap Kendrick, D. J . Organomet. Chem. 1991, 421, 241-246. Casellato, U.; Corain, B.; Graziani, R.; Longato, B.; Pilloni, G. Inorg. Chem. 1990, 29, 1193-1 198. Lonaato, B.:Pilloni.G.; Graziani, R.:Casellato,U. J . Oraanomet. Chem. 1991; 407,369-376. Kim, T.-J.; Kwon, S.-C.; Kim, Y.-H.; Heo, N. H.; Teeter, M. M.; Yamano, A. J. Organomet. Chem. 1991, 426, 71-86. Onaka, S.; Moriya, T.; Takagi, S.; Mizuno, A.; Furuta, H. Bull. Chem. SOC.. Jpn. 1992,65, 1415-1427. Phang, L.-T.; Au-Yeung, S. C. F.; Hor, T. S. A.; Khoo, S . B.; Zhou, Dalton Trans. 1993, 165-172. Z.-Y.; Mak, T. C. W. J . Chem. SOC., Cabri, W.; Candiani, I.; DeBernardinis, S.; Francalanci, F.; Penco, S.; Santo, R. J. Org. Chem. 1991, 56, 5796-5800. Madin, A.; Overman, L. E. Tetrahedron Lett. 1992, 33 (34), 48594862. Cramer, E.; Percec, V. J. Polym. Sci. Part A: Polym. Chem. 1990,28, 3029-3046. Galarini, R.: M u m , A.; Pontellini, R.; Bolognesi,A.; Destri, S.;Catellani, Chem. Commun. 1991, M.; Mascherpa, M.; Zhuo, G. J. Chem. SOC., 364-365. Trost, B. M.; Scanlan, T. S. J . Am. Chem. Soc. 1989,111,4988-4990. Carfagna, C.; Galarini, R.; Musco, A.; Santi, R. J. Mol. Catal. 1992, 72, 19-27. Brown, J. M.; Cooley, N. A. Organometallics 1990, 9, 353-359. Cabri, W.; Candiani, I.; Bedeschi, A.; Santi, R. J . Org. Chem. 1992,57, 35 58-3563. 1

important factor which contributes to the stabilization of the various metal geometries and the enhancement of catalytic activities of its complexes. This ligand flexibility imposes a large degree of complex fluxionality in solution which is a subject of this paper. A handful of semi-closo-bridged21 dppf complexes have been isolated and characterized. Single-crystal X-ray structure data for these complexes (Table 1) indicate that, despite the steric demand of the dppf ligand, it readily accommodates auxillary bridging ligands of smaller bite size through ring and/or torsional twisting. This unusual skeletal flexibility is clearly demonstrated for the complexes listed in Table 1. For these complexes, the deviation from ideal tetrahedral geometry about the Pi ( i = 1, 2; see Figure 1) is small for ai, where ai = LCi,(cp)-PrM containing the "ipso" in CiW(cp).In this article, C p is abbreviated to mean the substituted C5H4 rather than the conventional cyclopentadienyl C5H5. For [Re~(p-OMe)2(p-dppf)(CO)6]~~ and [ [Ag(NO3)(p~-dppf)]2.2CHCls],~~ this deviation is much larger because the significantly smaller bite angle of the auxillary bridges reinforced the steric rigidness of the bridging fragment. The rigidity of the latter complex may also be a result of the doublybridging framework. Our recent isolation of the dinuclear complex [Rez(p-0Me)2(pL-dppf)(C0)6]demonstrates unequivocally the ability for dppf to span two metal sites without a direct M-M link but which are rigidly held by two much smaller ligand bridges such as methoxyls.22 (1 8) Yoneyama, M.; Kakimoto, M.; Imai, Y. Macromolecules 1989,22,41524155. (19) Kalck, P.; Randrianalimanana, C.; Ridmy, M.; Thorez, A.; tom Dieck, H.; Ehlers, J. New J. Chem. 1988, 12, 679-686. (20) Knifton, J. F. J . Mol. Catal. 1988, 47, 99-116. (21) Semi-closo-bridged dppfcomplexes aredefined as di- (or poly-) nuclear complexes without a direct M-M bond but where the metal centers are bridged by dppf and other co-bridging ligands. (22) Yan, Y. K.;Chan, H. S. 0.;Hor, T. S. A.;Tan, K. L.; Liu, L. K.; Wen, Y. S.J . Chem. SOC.,Dalton Trans. 1992, 423-426. (23) Hor, A. T. S.;Neo, S. P.;Tan, C. S.; Mak, T. C. W.; Leung, K. W. P.; Wang, R. J. Inorg. Chem. 1992, 31, 45104516.

0020-166919411333-2407$04.50/0 0 1994 American Chemical Society

2408 Inorganic Chemistry, Vol. 33, No. 11, 1994

Lam et al.

Table 1. Summary of Phosphorus Bond Angles ai (a,= ~ C i ~ t c ~ l - Pfor ~ MQuasi-Closo ) Dppf-Bridged Complexes complexes

ai

113.7 112.0

Experimental Section All l H and 13C N M R spectra were recorded on a Bruker WM-250 superconducting FT-NMR spectrometer operating at 250.13 and 62.86 MHz, respectively. The chemical shift was reported in ppm units from internal TMS. For the variable-temperature ‘H N M R experiments, spectra were collected after the sample was equilibrated for 1 h. The COSY and EXSY experiments werecarried out using the standard COSY90 and NOESY sequence. The mixing time for the EXSY experiment was 0.4 s. For data treatment, a sine bell window function was used with a 0’ shift for both dimensions. Appropriate zero-filling was carried out in the second dimension prior to Fourier transformation. 13C spectral editing was carried out using standard DEPT sequence. All molecular modeling calculations were conducted using the CHEM-X program (PC release) licensed from the Chemical Design Ltd., Oxford, England. Davies et ~1.2’32 have successfully demonstrated that accessible conformational space can be generated through an analysis based on the assumption of correlated rigid rotor rotations. This type of analysis has proven to provide a reasonable explanation for observed conformations in solution. [Re2(r-OMe)z(r-dppfc06] was constructed using single-crystal X-ray structure data.22 Dummy atoms were used for Re and Fe because of the lack of adequate metal parametrization. Since steric influences from Fe atom were assumed to be negligible,29-32 contributions from steric effects to the potentials from Fe and Re atoms were neglected in the calculation. The van der Waals (vdW) interaction energy of the system sums the contribution arising from the electrostatic (Eelc),polarization (&I), and torsional (Em,) energy terms.” Their explicit forms are given as follows:

+

V,,, = 7’J 1.O cos 3w) V,,, = T,( 1.O - cos 2w) Figure 1. Labeling scheme for [Rez(r-oMe)2(r-dppf)(co)6].

Four interesting features are found in this complex: (a) a 1i= a2= 122.5O,whichis thelargestobservedforthisclassofcomplex.

(b) A closely related mononuclear complex [Re(OMe)(s2-dppf)(CO)3] is not formed despite the known stability of dppf as a chelating ligand and the apparent skeletal restriction of the bridging dppf on the [Rez(p-OMe)2] core. Recent isolation of proves the the analogous complex [Re(OMe)(q2-dppe)(CO)3]24 stability of this mononuclear configuration. (c) Fluxionality of the complex is reflected in, and confined to, the phenyl and Cp protons with the [Re2(pOMe)2] core remaining essentiallystatic. (d) The Cp proton N M R resonances spread over an unusually large region between 3.0 and 5.7 ppm. All these pecularities are related to the stereoproperties of dppf as a bridging ligand. This complex thus provides a suitable model for a detailed investigation of the solution dynamics of dppf in its semi-doso bridging state. The A-frame chemistry of alkyl-chained diphosphines especially dppm is rich and diverse.25 Similar dimetal chemistry of dppf is insofar undeveloped; its future prospect depends largely on the dynamic properties in relation to the metal cooperative effect of the ligand. A sample of the literature work revealed a host of -~ conformational fluxionality studies of chelating d p ~ f . ~In-depth analysis of bridging dppf is rare though recently the rocking motion of the dppf ligand between two metal centers has been recognized.2628 (24) Mandal, S. K.; Ho, D. M.; Orchin, M. Inorg. Chem. 1991, 30, 2 2 4 4 2248. (25) Chaudret, B.; Delavaux, B.; Poiblanc, R. Coord. Chem. Reu. 1988.86, 191-1 3 1. (26) Housecroft, C. E.; Rheingold, A. L. Organometallics 1987, 6, 13321340.

V,,, = Td(1.O - cos 2w)

nonconjugated bonds

(3)

conjugated bonds double bonds

In eq 1, Kis a unit conversion factor, QIand Q2 are the partial charges, tis the dielectric constant (or dielectric function), and r is the interatomic distance. In eq 2,A-Dareconstants, whereas ris theinternuclear distance in A. In eq 3, Tncis the nonconjugated barrier for the central atom pair, Tc is the conjugated barrier for the central atom pair, and Td is the double-bond barrier for the central atom pair, while w is the torsion angle.

Results and Discussion Assignments for the 1H and the 13C N M R spectra of [Re2(pOMe)2(pdppf)(C0)6] are summarized in Table 2. The downfield 13Cshift for the methoxy group (8(I3C) = 73.65 ppm) compared to that found in [Re(OMe)(dppe)(CO)3] (8(I3C) = 67.2 ppm22) indicates a reduction in the nucleophilicity of the oxygen upon conversion from terminal methoxo to its bridging state. FortheCpprotonsat 6(1H)= 3.64and4.78ppm,additional broadening was observed going from 60 to 250 MHz. Therefore, (27) Harpp, K. S.; Housecroft, C. E. J. Organomet. Chem. 1988,340,389396. (28) Draper,S. M.; Housecroft, C. E.; Rheingold,A. L. J . Organomet. Chem. 1992, 435, 9-20. (29) Davies, S. G.; Blackburn, B. K.; Whittaker, M. Stereochemistry of Organometallic and Inorganic Compounds; Bernal, I., Ed.; Elsevier: Amsterdam, 1989; Vol. 3, pp 141-223. (30) Davies, S . G.; Seeman, J. I.; Williams, I. H. Tetrahedron Lett. 1986, 27. . , 61 - - 9- 4-6-2- . (31) Davies, S . G.;Blackburn, B. K.; Sutton, H.; Whittaker, M. Chem. SOC. Rev. 1988, 17, 147-179. (32) Davies, S . G.; Derome, A. E.; McNally, J. P. J . Am. Chem. SOC.1991, 113, 28542861. (33) (a) CHEM-XReference Guide; Chemical Design Ltd.: London, 1993; Vol. 2, Chapter 14. (b) CHEM-X User Guide; Chemical Design Ltd.: London, 1992; Chapters 10 and 11.

Inorganic Chemistry, Vol. 33, No. 11, 1994 2409

Re2(ll-oMe)2(ll-dppf)(co)6

Figure 2. Suggested side and top view (a, left) a1.2 = 136' and (b, right) a1.2 = 142' of the structure of [Re2(p-OMe)2(p-dppf)(C0)6] passing from

R to the S forms. Table 3. Summary of AG* Values in CD2C12 for Cp Protons

Table 2. Summary of NMR Data in CDlCl2 Solution

_ _ _ _ _ _ _ _ _ _ ~

~~

w)/PPm 298 K 228 K 7.45 (bs)" 4.60 ( s ) ~ 3.64 (b)d 4.78 (b)d

7.05-7.38 4.60 3.00 (HS) 4.20 (Hb) 3.89 (Ha) 5.68 (Hd)

WC)/PPm

assgnt

73.6SC

phenyl protons p-CH3O group Cp protons Cp protons

Cp proton H, (3.00 ppm) Ha (3.89 ppm) Hb (4.20 PPm) Hd (5.68 PPm)

AG*/kJ mol-L 45.7 47.9 45.9 46.1 46.4 f 1.0 (mean)

Scheme 1

bs = broad singlet. s = singlet. This assignment is supported by DEPT experiments at 228 K. b = broad.

the fluxionality of Cp was examined since conformational exchange has been well established for metal complexes containing the dppf ligand system.24 The variable-temperature experimental results confirmed the fluxional properties of dppf. At 228 K in CD2C12,the exchanging Cp protons were frozen out giving four well-resolved peaks labeled H,, Ha, Hb, and Hd at 6(lH) = 3.00, 3.89,4.20, and 5.68 ppm, respectively. The phenyl absorptions were split into two sets of asymmetrical broad doublets. No changes in the methoxy peak was detected throughout the temperature-lowering process. This particular result indicates the following: (i) The structural integrity (rigidity) of the methoxy-bridged core ([r-OMe-Re]2) was retained in solution and was unaffected by the fluxional process. In the solid, the structural data demonstrated that the two methoxy ligands are symmetrically disposed between the two Re(1) centers with the averagedistanceofRe-0 = 2.163(6) AandRe-0-Re = 103.83(23)'. (ii) Little or no geometrical or electronic changes take place at the Re(1) centers as a result of the fluxional process. (iii) The fluxionality shown by the ferrocenyl fragment was unusually high. An average activation energy of 46.4 f 1.0 kJ/mol was calculated from the four peaks (Table 3) for the barrier to exchange. Fluxionality of the ferrocene similar to those in CD2C12 was also observed in CDCl3 and &-benzene. It was also observed that the phenyl rings remained dynamical in all three solvents at 195 K. During the freeze out process, peak cross-over occurred. Combined 2-D EXSY and 1HJH COSY experiments conducted at 228 K estabished that H,, Ha, Hb, and Hd all belong

to the same Cp and the two exchanging proton pairs are Ha, Hd and Hb, H,. The assignment for the C p protons is summarized in Table 2. The designation for the Cp protons and phenyl rings 4 and 4' is given in Figure 1. The expected J coupling is small (lJ = 1.5 Hz) compared to the line width of the protons -6 Hz; therefore, couplings were not resolved. Figure 1 was constructed using the CHEM-X program based on the publishedZ2single-crystal X-ray structure data of [Re2(~(-0Me)~(pdppf)(CO)6]. Two prominent features are noted: (i) 4 and 4' were oriented 21S0 with respect to each other and (ii) the C p F e - C p axis of dppf fragment is twisted by 53S0 with respect to the R w R e axis. If the "frozen" structure corresponds closely to that of the single-crystal X-ray structure, the ferrocene fragment of the dppf ligand in the complex adopts the synclinal34 (gauche) conformation. The relative orientation of the two Cp rings is w = 85.9' (+13.9O from an ideal gauche eclipsed conformation; see Figure 1). Therefore, the molecule has no symmetry plane. The conformers of [Rez(pOMe)z(p-dppf)(co)6] are enantiomeric with magnetically distinct phenyl groups. The pathway for the interchange of the R and S forms (Scheme (34) For a five-membered ring system, there are two idealized 'synclinal (or gauche)" conformations, vir. staggered ( w 36O) and eclipsed (w 72').

Lam et al.

2410 Inorganic Chemistry, Vol. 33, No. 11, 1994

8.5

48.5 aJ

2w U

v

-

88.5

Qi

IZX.5 135

140

145

1so

a (degrees) Figure 3. Plot of the relative vdW Energy versus a (deg) showing the angular dependence of the vdW energy with LCcp(ipso)-P-M. 1) can be visualized through the opposite twisting of the two halves of the ferrocene sandwich about the CpFe-Cp centroid axis coupled with the correlated torsional twist about the P-Ciwc,, and the P-Re bonds. The low-temperature 'H N M R spectrum confirmed the static [ABCDIzpattern for thecpprotons. Similar rocking motion has been recently advocated by Housecroft et al. on a [Au2M4B]( M = Fe,26327 Ru28) cluster core. For the conversion of the R to the S form, 4 and 4' must be identical during barrier crossing. This requirement is met when the Cp centroid-centroid axis (r,) is coplanar with the virtual P-P or the Re-Re axis. Although this structure can be obtained by rotation of both the P-Re and CQ,(~~~~)-P bonds, overlapping of the Cp electron cloud results within the approximation of the CPK35 model. Without increases in ai,r,, is greatly reduced below 3.28 A (centroid to centroid separation of the two cyclopentadienyl rings obtained from X-ray data) and the model is not meaningful. Marginal separation of the CPK clouds is obtained at a1.2 = 136" (Figure 2a). The corresponding rc< 2.75 A determined by measurement of the centroid to centroid distance remains significantly shorter than 3.28 A and the relative vdW is unrealistically high. Further increase of a1,2to -142" is necessary to avoid overlapping of the Cp ring cloud, while r,, is maintained at the optimized separation of 3.28 A (Figure 2b). The minimum on the plot of the relative vdW energy versus a (Figure 3) correspond to this structure. At a1,2 = 142", the corresponding r,, (-3.34 A) is in good agreement with the ideal r,, of 3.28 A. Therefore, both constraints were met simultaneously. The combined effects of tilting the Cp planes together with variation in a resulted in no significant improvements in the relative vdW energy, in the separation of the CPK cloud, or in the optimization of r,, toward 3.28 A. The large increase (=20" from 122") in a1,2 142" necessary to reach a vdW energy minimized structure during barrier crossing is questionable although some support can be derived from the values of the ai (> 109.S0) for similar semi-closo-bridged complexes collected in Table 1. We suggest that the transformation of the R to S forms is unlikely by passage over the barrier. Probably, quantum mechanical tunneling through the barrier is more plausible. To account for the large Cp proton shifts at 228 K, thedynamics of 4 and 4' in [Rez(c~-OMe)2(c~-dppf)(C0)6] were studied by following the well-established conformational analysis method set out by DavisZ9-3l using CHEM-X. The conformation calculations were restricted to intramolecular dynamic processes involving only 4 and 4f. All other groups bound to the phosphorus remain fixed because the rotation motions of the methoxy groups and the carbonyl groups are sufficiently far away from 4 or 4' to affect their dynamic properties. Therefore, they can be

-

-

(35) The CPK radius-hard sphere space-filling of individual atoms-was chosen for the study of the cloud effects.

168.5

'

I

I

205.4

155.4

I 105.4

0,(degrees)

175.4

165.4

155.4

145.4

135.4

0,(degrees) Figure 4. Conformationalenergy contour (kJ) diagram for &'(LC(26)-

C(2l)-P(l)-C(51) and@t(LC(l2)-C(Il)-P(l)-C(51)generatedusing CHEM-X. (a) Top: Full 180' range for both &' and & (b) Bottom: Expansion of conformation contours corresponding to the square (broken line) defined in (a). neglected from the conformational analysis. Four Cipo(phcny~)-P

bonds,namely,C(21)-P(1),C(41)-P(2),C(1l)-P(l),andC(31)P(2), were considered in the calculation. The conformational energy surface created by 4 and 4' was generated by arbitrarily setting, e.g. 4, in a fixed orientation and driving through 360". By repetition of this procedure for other angles of 4 in 5" increments, the resulting nonbonded conformation energy contours are shown in Figure 4a. The region defined by the square box (broken line) in Figure4a is expanded in Figure4b. All calculated nonbonded energies were rounded off to the first decimal place with the lowest value arbitrarily set to zero on the energy scale. In Figures 4a,b, counterclockwise rotation about the Cipso(phcnyl)-P bond for both 4 and 4' corresponds to increasing torsional angles defined by 4t = C(l2)-C(ll)-P(l)-C(51) and 4,' = C(26)C(21)-P(l)-C(51), respectively, and is illustrated in Figure 1. The two conformation wells are shown in grided areas labeled "A" and "B" in Figure 4a. The high-energy valley located below -167 kJ/mol (grided area "A") corresponds to the conformation in which the phenyl protons are sandwiched between the Cp rings (Figure 5 ) . The lowest energy pathway connecting "A" and "B" was determined to follow line "C" in Figure 4a. The highest point (barrier) along "C" was determined to be -400 kJ/mol. The remaining discussion focuses on "B" whereby all conformation states defined by the contour below 4 kJ/mol were fully accessible. #J~

Inorganic Chemistry, Vol. 33, No. 11, 1994 2411

P

?

Figure 5. Orientation of @ or @’ corresponding to grided area A in Figure 4a: (a, top) top view; (b, bottom) side view.

For the energy minimization procedure, the optimum orientations of dt and 4; were determined by initially setting the plane of the phenyl rings vertically relative to the C p ~ l a n e s . 2 The ~ torsional angles cPt and 4; equal 158.6 and 99.3’, respectively, in the energy-minimized structure. This structure marked by “M” in Figure 4b corresponds to the conformation generated by 4 and 4’ as ideal rotors. In Figure 4b, the contour at 4 kJ/mol defines the boundary for the rotation motion of the Ci,,-P bond for 4 and 4’. The rotational limits, 92.6’ I4; I112.4’, was determined for 4’. Similarly, 4 was determined to oscillate between the limits 147.6’ I4t I177.8’. Rotation of 4 and 4’ is determined to be opposite for strain relief because of ring cloud repulsion. On the basis of the rotation limits of Cipso(+,+~)-P in mobile solution at 228 K, the average torsional angles were obtained, i.e. 41 = 162.7’ and &’ 102.5’. The average conformation determined by this procedure is located at the center of the potential well marked by the point “S”in Figure 4b. In the crystal form, +t = 155.4’ was determined for 4 and 4,‘ = 108.5’ for 4’. This point is marked by “X” in Figure 4b. This conformation was = l kJ/mol higher than the lowest energy conformation generated. The difference in the torsional angles determined between solid and solution was small and is attributed to packing effects. Therefore, the X-ray structure is confirmed to be thermodynamically the lowest energy conformation. Given the large number of approximations inherent in the use of modeling techniques, the agreement between the values of the torsional angles using the procedures described above compared to the values determined by single-crystal X-ray is excellent. The triangle in Figure 4b defined by the three points “X”, “S”,and “M” bracketed the true values for 4t and 4; in solution. The distances of the respective Cp protons from 4 or 4’ were readily obtained from the torsional rotation limits for the CiW(+,+))-P bonds. From the spatial orientation of the Cp protons at the torsional limits for 4’ (Figure 6), it was determined that neither H, nor Hb was in the proximity of either 4 or 4’. Consequently, normal C p proton chemical shifts were expected and were confirmed experimentally. The average distance of H, was determined to be =2.86 8, above the ?r-cloud of 4’. Aside from ring current effects which led to shift of H, to high field,

Figure 6. Structural representations showing the spatial orientation of the Cp protons and the intramolecular interligand H,,d(Cp)-f#J’(plane) distanceswith respect to the three rotational limits for the Ci,-P bonds.

: ________________________________________-Ordinary Phenyl Proton Chemical Shifts ..........................................

T -

2

2.2

2.6

2.4

2.8

3

(‘4 Figure 7. Plot of &‘Hi)versus r for cyclophanes. Data were tabulated in Table 4. Key: (1) Compound E, Ha; (2) compound B; (3) compound D, Hb; (4) compound E, Hb; ( 5 ) compound A; (6) compound D, Ha; (7) compound D, H,; (8) compound F; (9) compound C.

there is no spectroscopic evidence suggesting H, interacts with the ?r electron cloud of the phenyl ring in the form of CH/?r i n t e r a c t i ~ n . The ~ ~ . magnitude ~~ in the upfield shift for H, is similar ( 3 6 ) Nishio, M.; Hirota, M. Tetrahedron 1989, 45, 7201-7245.

2412 Inorganic Chemistry, Vol. 33, No. 11, 1994

Lam et al.

Table 4. Summary of IH N M R Data for Cyclophanes"

cyclophanes were constructed using data38 from solid-state and/ or standard bond lengths and angles using CHEM-X. The distance r was measured when the energy minimized structure was obtained. For ordinary phenyl protons, 6(IH) i= 6.8-7.5 ppm (Figure 7). When r = 3 A, as much as 1.5 ppm shift to upfield could be expected for normal phenyl protons. Therefore, the observed upfield shift of 1 ppm from =4 ppm (normal Cp proton chemical shifts) for H, (6(lH) = 3.00 ppm) is readily rationalized on the basis of the analysis built upon the internal proton (Hi) chemical shifts observed for cyclophanes. The rotation restriction of 4' at 228 K ensures that Hd, always sitting in the quasi-closo-dibridged cavity, maintains a position within very close proximity to the deshielding region of 4'. The average distance of Hd from the nearest carbon atom on 4' was determined to be 2.95 A. Again, ring current effects accounts for the large proton shift to low field. Downfield shift of one of the a-protons, Hd, has also been observed in the [HRud(C0)12BAu2(dppf)] cluster (6(lH) = 5.37 ppm2*) when it is pointing toward the [ A u ~ P cavity. ~] It is anticipated that ring current effects are also operative. The observed shift in the present complex (6(lH) = 5.68 ppm) is to our knowledge the largest shift for reported Cp resonances.

4.25

2.34

3.72

2.21

4.41

2.52

3.85 (Ha) 3.59 (Hb) 4.21 (H,)

2.35 2.20 2.39

Conclusion

3.67 (Ha) 3.57 (Hb)

2.10 2.21

3.94

2.49

Reference 38.

to the upfield shift phenomenon observed for the internal proton, Hi, in cyclophane molecules. Ring current effects has been established to be the origin responsible for the shifts of Hi when situated above the a-cloud. The Hi proton chemical shifts depends on its distance r (the perpendicular distance between the proton and the ring plane) above the a-cloud; the shorter the distance, the more pronounced the upfield shift. Figure 7 gives the plot of &(]Hi)versus r for the cyclophanes collected in Table 4. The (37) Sakakibara, K.; Hirota, M. Chem Lett. 1989, 921-924.

Previous works have already identified the rotational motions of the Cp rings, inversion and vertical movements of the phosphorus atoms, and twisting motion of the ferrocenyl skeleton with respect to the dimetal entitie~.~-IOThe present work enables an illustration of these fluxionalities through an N M R conformational analysis. It also for the first time reveals the spectroscopic and geometrical impact of the phenyl rotations. These rotations, in conjunction with the ferrocenyl mobilities, allow the dppf ligand to seek and adopt a most suitable conformation between two metal centers. That such a comfortable stereoposition can be found irrespective of the metal environment and other complex core behavior is a vital prerequisite for this ligand to function effectively in a changeable role. The stability of the dppf ligand on the rigid [Rez(pOMe)2] core with respect to dissociation to give the mononuclear chelate form is a consequence of this ligand adaptability. An important lead of this work lies in the analysis of other diphosphine complexes whose catalytic activities are believed to be conformationally related. The advantage of the dppf ligand as an active probe for N M R analysis of phosphine complexes can be envisaged. Supplementary Material Available: Figures showing IH N M R , 2D EXSY, and 2D COSY-90 spectra (3 pages). Ordering information is given on any current masthead page. (38) Keehn, P. H., Rosenfeld, S. M., Eds. Cyclophanes; Academic Press: New York, 1983; Vols. I and 11.